CN104137568B - Frequency characteristic transformation device - Google Patents

Abstract

It is equipped with: HPF (702), which deforms the frequency characteristic of the target signal; a phase correction unit (701), which corrects the phase characteristic of the target signal, and makes it approximately the same as the phase characteristic of the HPF (702); the first multiplication The device (705) adjusts the gain of the signal output from the phase correction unit (701); the second multiplier (706) adjusts the gain of the signal output from the HPF (702); the coefficient determination unit determines the first and second The gain coefficient of the second multiplier (705, 706) is such that the sum of the gain coefficient of the first multiplier (705) and the gain coefficient of the second multiplier (706) becomes a fixed value; The two signals output by the first multiplier (705) and the second multiplier (706) are summed.

Description

Frequency characteristic deformation device

technical field

The invention relates to a signal processing technology for improving distortion and broken sound in sound signal reproduction.

Background technique

In a speaker reproduction system that reproduces sound signals such as music and broadcasts by a speaker, the sound quality may deteriorate due to distortion or broken sound. There are roughly two causes of distortion and broken sound. The first is when the input signal to the speaker is distorted, and the second is when the input signal is not distorted, but the reproduction limit of the speaker is exceeded and distortion or broken sound occurs.

The first case can be explained as follows. In recent sound signal reproduction systems, devices for correcting frequency characteristics or adjusting volume by digital processing are increasing. In the correction of the frequency characteristic, for example, if the high-frequency component is increased by 10 dB, the digital signal may be saturated at a volume value of -10 dBFS or higher. In addition, 0dBFS represents the maximum amplitude value of a digital signal. Therefore, the reproduced sound is digitally distorted at high volume, resulting in deterioration of sound quality. Figure 2 shows the situation.

In FIG. 2 , the vertical axis represents the amplitude strength of the digital signal, and the horizontal axis represents the frequency. In addition, the region where the signal is saturated and the sound breaks occurs is shown in gray, and its boundary is shown in thick lines. 201, 202, and 203 denote an example of frequency characteristics of a digital audio signal whose frequency characteristics have been corrected, 201 is a characteristic of a low volume value, 202 is a characteristic of a medium volume value, and 203 is a characteristic of a large volume value. In the volume values of 201 and 202, the sound signal does not exceed 0dBFS, so that the sound does not break and can be enjoyed with the original sound quality. However, if the volume is raised as in 203, the signal strength of a part of the high-frequency components exceeds 0 dBFS and becomes digitally saturated. If the signal is saturated, distortion occurs, the sound breaks and the sound quality deteriorates.

In this way, if a digital signal whose frequency characteristic is corrected is to be reproduced at a large volume, a specific frequency component may exceed 0dBFS which is the maximum amplitude value of the digital signal, so that the signal is saturated to cause distortion and broken sound.

Next, the second case, that is, the case where distortion and broken sound occurs due to exceeding the reproduction limit of the speaker, will be described.

In speaker reproduction, there is a maximum displacement range in which the speaker diaphragm can vibrate. If a signal exceeding the maximum displacement range is input, the speaker diaphragm cannot vibrate satisfactorily, resulting in distortion and broken sound. Here, the magnitude of displacement of the speaker diaphragm depends on the frequency of the input signal. Figure 3 shows the relationship. FIG. 3 shows the displacement width of the speaker diaphragm when only the frequency of the signal is changed without changing the voltage (V) and input to the speaker. In addition, in FIG. 3 , the characteristics near F0 are more convex or smoother than those in FIG. 3 due to the difference in the Q value indicating the degree of braking of the speaker, etc., but the general tendency does not change. In addition, since the present invention can also be applied to a speaker whose displacement width characteristic is different from that shown in FIG. 3 , for convenience of description, the displacement width characteristic of the speaker diaphragm will be described as FIG. 3 .

In FIG. 3 , the displacement width of the speaker diaphragm is approximately constant at frequency components lower than F0 (lowest resonance frequency of the speaker), and the displacement width is at a slope of approximately -12dB/oct at frequency components higher than F0. And reduce. This means that when a low frequency component below the vicinity of F0 is input to the speaker, the speaker diaphragm vibrates with a larger displacement width than when a high frequency component is input. Therefore, if a signal containing a large number of low frequency components is input to the speaker and the voltage thereof is increased, the maximum displacement width of the diaphragm will be exceeded above a certain voltage. That is, it can be said that the more the signal contains a large number of low frequency components, and the higher the voltage is, the easier it is to exceed the reproduction limit of the speaker. Figure 4 shows the situation.

In FIG. 4 , the vertical axis represents the amplitude strength of the signal, and the horizontal axis represents the frequency. In addition, the region where the sound breaks beyond the displacement limit of the speaker diaphragm is shown in gray, and its boundary is shown in thick lines. Here, the characteristic of FIG. 4 is the characteristic of the amplitude value of the audio signal, so unlike the characteristic of the displacement width of the speaker shown in FIG. 3 , the displacement limit of the speaker diaphragm has a slope of +12dB/oct.

In addition, 401, 402, and 403 indicate the frequency characteristics of the audio signal reproduced by the speaker, and it is assumed that a large number of low-frequency components are contained in particular. 401 is the characteristic when the volume value is small, 402 is the characteristic when the volume value is medium, and 403 is the frequency characteristic when the volume value is large. In the part reproduced at a small volume value like 401, even if the sound signal contains a large amount of low-frequency components, it will not exceed the maximum displacement range of the speaker diaphragm, so the sound can be played with the original sound quality without breaking the sound. enjoy. However, if the volume is increased as in 402 and 403, the maximum displacement range of the speaker diaphragm will be exceeded, so distortion and broken sound will occur and the sound quality will deteriorate.

In this way, if a signal exceeding the maximum displacement width of the diaphragm is input, the diaphragm cannot vibrate satisfactorily, and distortion and broken sound occur.

Distortion and broken sound are sounds that are not included in the original sound signal, and thus become a major factor that hinders enjoyment of music.

With regard to this problem, the processing structure shown in FIG. 17 has conventionally alleviated distortion and broken sound. In FIG. 17 , an input signal 1301 is configured to output a signal 1303 after passing through an HPF (High Pass Filter) 1302 that suppresses low-frequency components. With such a configuration, it is possible to suppress the low-frequency components that cause crackling before being input to the speaker, so that the rate of occurrence of distortion and crackling can be reduced. However, in the conventional technology, there is a problem that, since the low-frequency components are suppressed by the HPF1302, even when, for example, the reproduced signal has few low-frequency components and the speaker is driven with a large voltage, there is always a problem. The low-frequency components are suppressed and the original sound cannot be reproduced. In addition, there is another problem that the low frequency components are always suppressed even when the high-pass filter 1302 is not used for driving without the high-pass filter 1302, so that the original sound cannot be reproduced. That is, in the conventional technology, there is a problem that, in order to prevent sound cracking when driving with a large voltage (during loud volume), the low-frequency component is excessively suppressed, and the original sound quality cannot be enjoyed.

As a technique for alleviating this problem, there is a technique disclosed in Patent Document 1, for example. FIG. 18 is a processing block of the amplitude limiting device disclosed in Patent Document 1. FIG. According to Patent Document 1, in the amplitude limitation for suppressing excessive input, the amount of distortion due to the amplitude limitation characteristic is detected, and the gain of each frequency band is controlled according to the value, thereby alleviating the sound quality deterioration due to the amplitude limitation Difference.

Patent Document 1: Japanese Patent Laid-Open No. 2009-147701

Contents of the invention

However, in the technology disclosed in Patent Document 1, the frequency components to be suppressed are limited to the divided frequency bandwidth, and therefore there is a problem that even frequency components that should not be suppressed are excessively suppressed and the sound quality deteriorates. For example, consider a case where the frequency bandwidth divided into subbands by a BPF (Band Pass Filter) is 100 Hz. At this time, if a signal having a large intensity in the frequency components below 60 Hz is input, only the signal components below 60 Hz are originally suppressed so that no crackling occurs. However, in this technique, the intensity of all signal components of 0 to 100 Hz is suppressed, so components other than the frequency components to be suppressed (components of 60 to 100 Hz) are also suppressed. Also, as shown in FIG. 3 , the displacement width of the speaker amplitude plate has a frequency characteristic, but the amplitude limiting device disclosed in Patent Document 1 does not have a processing structure reflecting the frequency characteristic of the displacement width. Therefore, it can be said that it does not have the function of preventing sound breakage caused by exceeding the maximum displacement of the speaker diaphragm.

The present invention was made to solve the above-mentioned problems, and an object of the present invention is to provide a frequency characteristic deforming device capable of preventing distortion and broken sound during speaker reproduction while maintaining sound quality.

The frequency characteristic deforming device of the present invention includes: a filter that deforms the frequency characteristic of the target signal; a phase correcting unit that corrects the phase characteristic of the target signal so that it is substantially the same as the phase characteristic of the filter; A multiplier, which adjusts the gain of the signal output from the phase correction unit; the second multiplier, which adjusts the gain of the signal output from the filter; the coefficient determination unit, which determines the gain coefficients of the first and second multipliers so that the first multiplier The sum of the gain factor of the multiplier and the gain factor of the second multiplier becomes a fixed value; and the adder adds the two signals output from the first multiplier and the second multiplier.

According to the present invention, it is possible to provide a frequency characteristic deforming device that prevents distortion and broken sound during speaker reproduction while maintaining sound quality.

Description of drawings

FIG. 1 is an explanatory diagram of the principle of a frequency characteristic deformation device according to Embodiment 1. FIG.

FIG. 2 is a graph showing the relationship between the amplitude limit of a digital signal and the frequency characteristic of a sound source.

FIG. 3 is a graph showing displacement characteristics of a speaker diaphragm.

FIG. 4 is a graph showing the relationship between the vibration limit of the speaker and the frequency characteristic of the sound source.

5 is an explanatory diagram showing changes in frequency characteristics due to two gains of the frequency characteristic deforming device according to Embodiment 1. FIG.

FIG. 6 is an explanatory diagram of the principle of a frequency characteristic deformation device according to Embodiment 2. FIG.

FIG. 7 is an explanatory diagram showing changes in frequency characteristics due to two gains in the frequency characteristic deformation device according to Embodiment 2. FIG.

FIG. 8 is an explanatory diagram of the principle of a frequency characteristic deformation device according to Embodiment 3. FIG.

FIG. 9 is an explanatory diagram showing changes in frequency characteristics due to three gains in the frequency characteristic deformation device according to Embodiment 3. FIG.

FIG. 10 is an explanatory diagram of the principle of a frequency characteristic deformation device according to Embodiment 4. FIG.

FIG. 11 is a diagram showing an example of a low-frequency extraction unit of the frequency characteristic deformation device according to Embodiment 4. FIG.

FIG. 12 is a diagram showing another example of a low-frequency extraction unit of the frequency characteristic deformation device according to Embodiment 4. FIG.

FIG. 13 is a diagram showing an added image of harmonics corresponding to the low-frequency attenuation effect of Embodiment 4. FIG.

FIG. 14 is an explanatory diagram of the principle of a frequency characteristic deformation device according to Embodiment 5. FIG.

FIG. 15 is a diagram showing an added image of harmonics corresponding to the low-frequency attenuation effect of Embodiment 5. FIG.

FIG. 16 is an explanatory diagram of the principle of a frequency characteristic deformation device according to Embodiment 6. FIG.

Fig. 17 is an explanatory diagram of the principle of the prior art.

Fig. 18 is a processing block diagram of a conventional amplitude limiting device.

(Symbol Description)

101: input signal; 102: excessive input estimation unit; 103: frequency characteristic deformation unit; 105: control unit; 107: output signal; 501: speaker diaphragm displacement estimation unit; 502: information; 701: phase correction unit; 702 : HPF; 705: first multiplier; 706: second multiplier; 713, 2003, 2125, 2126, 2127, 2204: adder; 901, 2012: LPF; 1101: first HPF; 1102: second HPF; 1103: third HPF; 1107: first phase correction unit; 1108: second phase correction unit; 1109: third phase correction unit; 1113: first multiplier; 1114: second multiplier; 1115: third multiplier ;2001: High-order harmonic signal generation part; 2004: Low-frequency extraction part; 2006: High-order harmonic generation part; 2008, 2119, 2120, 2121, 2201: Multiplier; 2011: Differential device; 2101: The first LPF; 2102: second LPF; 2103: third LPF; 2107: fourth phase correction unit; 2108: fifth phase correction unit; 2109: sixth phase correction unit; 2113: first higher harmonic generation unit; The second high-order harmonic generation department; 2115: the third high-order harmonic generation department.

detailed description

Hereinafter, in order to explain this invention in more detail, the form for implementing this invention is demonstrated based on drawing.

Implementation mode 1.

FIG. 1 is a diagram showing an embodiment of the present invention. Next, the operation of this embodiment will be described.

The input signal 101 input to the frequency characteristic deformer of the present invention is branched and sent to the phase correction unit 701 and the HPF 702 .

The phase correction unit 701 does not change the frequency-amplitude characteristic of the input signal, but only corrects the phase characteristic so that it becomes substantially the same as the phase characteristic of the HPF 702, and outputs the obtained signal 703 to the first multiplier 705 and the excessive input estimation unit. 102.

The HPF 702 filters the input signal 101 and outputs the obtained signal 704 to the second multiplier 706 and the excessive input estimation unit 102 .

Here, an implementation method of the phase correcting unit 701 that corrects the phase so as to obtain substantially the same phase characteristic as that of the HPF 702 will be described. When the HPF702 is realized with a 1st-stage 2nd-order IIR filter, its phase characteristic rotates exactly 90 degrees at the cutoff frequency, and gradually rotates to 180 degrees in the subsequent frequency components. The phase correction unit 701 that realizes such a phase characteristic can be constituted by an all-pass filter using a first-order IIR filter. In addition, when HPF is realized with two stages of second-order IIR filters, the phase characteristic is rotated exactly 180 degrees at the cutoff frequency, and gradually rotates to 360 degrees in the frequency components after that. The phase correction unit realizing such a phase characteristic can be constituted by an all-pass filter using a second-order IIR filter. In addition, in the case of realizing HPF with N stages of second-order IIR filters, the same phase characteristics can be realized by appropriately connecting in series first-order IIR and second-order IIR all-pass filters. In addition, when the HPF is realized by using an FIR filter, the phase characteristic becomes a linear phase, so the phase correction unit 701 can be configured by sampling delay processing. In this way, the phase correction unit 701 having the same phase characteristics as the HPF 702 can be realized.

The excessive input estimation unit 102 of this embodiment is constituted by a speaker diaphragm displacement estimation unit 501 . The speaker diaphragm displacement estimation unit 501 estimates the speaker diaphragm displacement value when the signal 703 is reproduced using information 502 such as the volume value and F0 of the target speaker, and obtains the first speaker diaphragm displacement value 707 . Similarly, the displacement value of the speaker diaphragm when the signal 704 is reproduced is estimated, and the second speaker diaphragm displacement value 708 is obtained. As a specific example of displacement value estimation, an LPF using a second-order IIR filter with F0 as the cutoff frequency is prepared, and the input signal is passed through the LPF and then multiplied by the volume value to obtain a value approximately proportional to the displacement width of the target speaker. value. In addition, since the Q value can also be changed in the LPF using the second-order IIR filter, it is also possible to improve the estimation accuracy by changing the Q value according to the braking degree of the target speaker. Of course, other methods, such as FIR filters, can also be used to simulate the vibration plate displacement characteristics of the target speaker.

The two speaker diaphragm displacement values 707 and 708 obtained in this way are output to the control unit 105 .

In the control unit 105, when the two input speaker diaphragm displacement values 707 and 708 are multiplied by different gain coefficients and then added, the absolute value of the amplitude value converges within a predetermined threshold value. a gain factor. However, the total of the two gain coefficients is set to 1.

If the two gain coefficients are changed under such conditions, different low-frequency attenuation effects can be achieved. FIG. 5 shows frequencies based on two gain coefficients when the HPF 702 is realized by a two-stage second-order IIR filter with a cutoff frequency of 80 Hz, and the phase correction unit 701 is realized by a one-stage second-order IIR all-pass filter with a cutoff frequency of 80 Hz. Changes in characteristics. In addition, in FIG. 5 , when the gain coefficient for the speaker diaphragm displacement value 707 is A1 and the gain coefficient for the speaker diaphragm displacement value 708 is A2, 301 indicates the characteristics of A1=1.0 and A2=0.0 , 302 represents the characteristic of A1=0.1, A2=0.9, and 303 represents the characteristic of A1=0.0, A2=1.0. Thus, it can be seen that from the completely flat characteristics (A1=1.0, A2=0.0), it is possible to achieve Different low frequency attenuation characteristics. In addition, since frequency components equal to or greater than the cutoff frequency are added in such a way that the total becomes 1, the components for matching the phases can be maintained, so that the intensity does not increase or decrease, and flat characteristics can be maintained.

As a specific calculation method of such two gain coefficients, if the speaker diaphragm displacement value 707 is set to X1, the speaker diaphragm displacement value 708 is set to X2, the gain coefficient for X1 is set to A1, and the gain coefficient for X2 is set to When the gain coefficient is A2 and the predetermined threshold is T, it can be realized by obtaining A1 and A2 satisfying the following formula (1).

T>ABS(X1×A1+X2×A2)···(1)

A1+A2=1

In addition, ABS(x) represents the absolute value of x.

In addition, in order to suppress the distortion of the frequency characteristic to the necessary minimum, it is preferable to find a combination in which the value of A1 is close to 1 among the combinations of A1 and A2 satisfying the above-mentioned formula (1). This is because A1 is a signal based on a signal obtained by correcting only the phase characteristic, and the closer A1 is to 1, the less distortion of the frequency characteristic. In order to obtain such a gain coefficient, first, set A1=1, obtain the value of ABS (X1×A1+X2×A2) while gradually reducing the value of A1, and use A1 and A2 at a time smaller than T That's it.

A1 obtained in this way is output to the first multiplier 705 as a gain coefficient 709 . In addition, A2 is output to the second multiplier 706 as the gain coefficient 710 .

In the first multiplier 705 , the input signal 703 is multiplied by the gain coefficient 709 , and the obtained signal 711 is output to the adder 713 .

In the second multiplier 706 , the input signal 704 is multiplied by the gain coefficient 710 , and the obtained signal 712 is output to the adder 713 .

The adder 713 adds the two input signals 711 and 712 and outputs the obtained signal as the output signal 107 .

As described above, with the processing configuration of the first embodiment, it is possible to prevent the reproduced audio signal from being excessively input. Therefore, according to the present invention, it is possible to obtain the effect that distortion and broken sound can be suppressed. In addition, by lowering the cutoff frequency as much as possible by the control unit, it is also possible to obtain the effect that distortion and broken sound can be prevented with the necessary minimum frequency characteristic change.

Implementation mode 2.

By substituting the HPF 702 described in Embodiment 1 with an LPF, it is possible to adjust the frequency of the signal so that the frequency of the signal does not exceed the maximum amplitude of the digital signal in a digital audio signal that has undergone frequency characteristic correction such that the high-frequency component is much corrected. Character deformation. FIG. 6 shows the processing structure of an example when the HPF 702 is replaced with an LPF.

The input signal 101 input to the frequency characteristic deformer of the present invention is branched and sent to the phase correction unit 701 and the LPF 901 .

The phase correction unit 701 does not change the frequency-amplitude characteristic of the input signal, but only corrects the phase characteristic so that it becomes approximately the same as the phase characteristic of the LPF 901, and outputs the obtained signal 703 to the first multiplier 705 and the excessive input estimation unit. 102.

The LPF 901 filters the input signal 101 and outputs the obtained signal 902 to the second multiplier 706 and the excessive input estimation unit 102 .

Here, the phase correction unit 701 can be realized by an all-pass filter or sampling delay processing similarly to the first embodiment, and thus detailed description thereof will be omitted.

The excessive input estimation unit 102 of this embodiment is composed of a digital signal amplitude calculation unit 601 . In the digital signal amplitude calculation unit 601 , the volume value 602 is multiplied by the input signal 703 to obtain a first amplitude value 707 . Similarly, the second amplitude value 708 is obtained by multiplying the volume value 602 and the input signal 902 .

The two amplitude values 707 and 708 obtained in this way are output to the control unit 105 .

In the control unit 105, when the two input amplitude values 707 and 708 are multiplied by two different gain coefficients and added together, the two amplitude values whose absolute value converges within a predetermined threshold are obtained. gain factor. However, the total of the two gain coefficients is set to 1. In addition, the predetermined threshold generally refers to setting 0 dBFS, but it is not limited to this, and a smaller value may be set when it is desired to limit the output of the amplifier without matching the input impedance of the speaker and the output of the amplifier.

If the two gain coefficients are changed under such conditions, different high-frequency attenuation effects can be realized. FIG. 7 shows two gain coefficients when the LPF 702 is realized by a two-stage second-order IIR filter with a cutoff frequency of 6000 Hz, and the phase correction unit 701 is realized by a first-stage second-order IIR all-pass filter with a cutoff frequency of 6000 Hz. changes in frequency characteristics. In addition, in FIG. 7 , when the gain coefficient for the speaker diaphragm displacement value 707 is A1 and the gain coefficient for the speaker diaphragm displacement value 708 is A2, 801 indicates the characteristics of A1=1.0 and A2=0.0 , 802 represents the characteristic of A1=0.1, A2=0.9, and 803 represents the characteristic of A1=0.0, A2=1.0. In this way, it can be seen that different high-frequency attenuation characteristics can be realized from completely flat characteristics (A1=1.0, A2=0.0) to the characteristics (A1=0.0, A2=1.0) of the two-stage 2nd-order IIR filter with a cutoff frequency of 6000 Hz . In addition, since frequency components equal to or less than the cutoff frequency are added at a rate of 1 in total, the components for matching the phases can be maintained, so that the intensity does not increase or decrease, and flat characteristics can be maintained. As a specific calculation method of such two gain coefficients, it is obtained in the same manner as in the first embodiment, and therefore description thereof is omitted.

A1 obtained in this way is output to the first multiplier 705 as the gain coefficient 709 . In addition, A2 is output to the second multiplier 706 as the gain coefficient 710 .

In the first multiplier 705 , the input signal 703 is multiplied by the gain coefficient 709 , and the obtained signal 711 is output to the adder 713 .

In the second multiplier 706 , the input signal 902 is multiplied by the gain coefficient 710 , and the obtained signal 712 is output to the adder 713 .

In the adder 713 , the two input signals 711 and 712 are added, and the obtained signal is output as the output signal 107 .

As described above, according to the processing structure of the second embodiment, it is possible to suppress correction of the amplitude value of the digital audio signal such as a high-frequency component much, and it is possible to obtain an effect that distortion and crackling can be suppressed. In addition, in the control unit of this embodiment, the cutoff frequency of the LPF can be made as high as possible, so that distortion and broken sound can be prevented with the minimum necessary frequency characteristic change.

Implementation mode 3.

In Embodiments 1 and 2, the frequency characteristic deformation part is realized by one phase correction part and one HPF or LPF, but it is not limited to this, and the frequency characteristic deformation can also be realized by multiple phase correction parts and multiple HPFs or LPFs department.

FIG. 8 is a diagram showing an example in which a frequency characteristic deformer is realized by three phase correctors and three HPFs. Next, the operation of this embodiment will be described.

The input signal 101 input to the frequency characteristic deformer of the present invention is branched into three, and sent to the first HPF1101, the second HPF1102, and the third HPF1103.

In the first HPF 1101 , the input signal is filtered, and the obtained signal 1104 is output to the first phase correction unit 1107 .

In the second HPF 1102 , the input signal is filtered, and the obtained signal 1105 is output to the second phase correction unit 1108 .

In the third HPF 1103 , the input signal is filtered, and the obtained signal 1106 is output to the third phase correction unit 1109 .

In the first phase correction unit 1107, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the second HPF 1102 and the third HPF 1103 are processed. The obtained signal 1110 is output to the first multiplier 1113 and the excessive input estimation unit 102 .

In the second phase correction unit 1108, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the first HPF 1101 and the third HPF 1103 are processed. The obtained signal 1111 is output to the second multiplier 1114 and the excessive input estimation unit 102 .

In the third phase correction unit 1109, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the first HPF 1101 and the second HPF 1102 are processed. The obtained signal 1112 is output to the third multiplier 1115 and the excessive input estimation unit 102 .

Here, each phase correction unit can be realized by an all-pass filter or sampling delay processing similarly to the first embodiment, and thus detailed description thereof will be omitted.

The excessive input estimation unit 102 of this embodiment is constituted by a speaker diaphragm displacement estimation unit 501 . The speaker diaphragm displacement estimation unit 501 estimates the speaker diaphragm displacement value when the signal 1110 is reproduced using the information 502 such as the volume value and F0 of the target speaker, and obtains the first speaker diaphragm displacement value 1116 . Similarly, the displacement value of the speaker diaphragm when the signal 1111 is reproduced is estimated, and the second speaker diaphragm displacement value 1117 is obtained. Similarly, the displacement value of the speaker diaphragm when the signal 1112 is reproduced is estimated, and the third speaker diaphragm displacement value 1118 is obtained.

As a specific example of the displacement value estimation, it is obtained by the same method as in the first embodiment, so detailed description is omitted.

The three speaker diaphragm displacement values 1116 , 1117 , and 1118 obtained in this way are output to the control unit 105 .

In the control unit 105, when the three input speaker diaphragm displacement values 1116, 1117, and 1118 are multiplied by different gain coefficients and added, the absolute value of the amplitude value converges within a predetermined threshold value. The 3 gain coefficients. However, the total of the three gain coefficients is set to 1.

If the three gain coefficients are changed under such conditions, different low-frequency attenuation effects can be achieved. Fig. 9 shows the change of the frequency characteristic based on 3 gain coefficients in the following situation, wherein, the situation is: the first HPF1101 is realized with a 2-stage 2-order IIR filter with a cutoff frequency of 30 Hz, and the first HPF 1101 is realized with a 2-stage IIR filter with a cutoff frequency of 70 Hz. The second HPF1102 is realized by a 2nd-order IIR filter with a cut-off frequency of 140Hz, and the third HPF1103 is realized with a 4-stage 2-order IIR filter with a cutoff frequency of 140Hz. Two stages of second-order IIR filters are connected in series to realize the first phase correction unit 1107, and a first-stage second-order IIR filter with a cutoff frequency of 30 Hz and a second-stage second-order IIR filter with a cutoff frequency of 140 Hz are connected in series. The second phase correction unit 1108 is realized by combining them, and the third phase correction unit 1109 is realized by connecting in series a single-stage second-order IIR filter with a cutoff frequency of 30 Hz and a single-stage second-order IIR filter with a cutoff frequency of 70 Hz.

In addition, in FIG. 9 , when the gain coefficient for the speaker diaphragm displacement value 1116 is set to A1, the gain coefficient for the speaker diaphragm displacement value 1117 is set to A2, and the gain coefficient for the speaker diaphragm displacement value 1118 is set to When it is A3, 901 represents the characteristics of A1=1.0, A2=0.0, A3=0.0, 902 represents the characteristics of A1=0.1, A2=0.9, A3=0.0, and 903 represents the characteristics of A1=0.0, A2=1.0, A3=0.0 The characteristic, 904 represents the characteristic of A1=0.0, A2=0.1, A3=0.9, 905 represents the characteristic of A1=0.0, A2=0.0, A3=1.0. In this way, it can be seen that the characteristics (A1 = 1.0, A2 = 0.0, A3 = 0.0) of the 2-stage IIR filter with a cutoff frequency of 30 Hz to the characteristics of a 4-stage second-order IIR filter with a cutoff frequency of 140 Hz (A1 = 0.0 , A2=0.0, A3=1.0), different low-frequency attenuation characteristics can be realized. In addition, frequency components equal to or greater than the cutoff frequency are added in such a way that the sum of them becomes 1 that the phases are aligned. Therefore, flat characteristics can be maintained without increasing or decreasing the intensity.

In addition, as a specific calculation method of such three gain coefficients, assuming that the speaker diaphragm displacement value 1116 is X1, the speaker diaphragm displacement value 1117 is X2, and the speaker diaphragm displacement value 1118 is X3, then The gain factor for X1 is A1, the gain factor for X2 is A2, the gain factor for X3 is A3, and the predetermined threshold is T, then it can be obtained by obtaining the following formula (2): A1, A2, A3 to achieve.

T>ABS(X1×A1+X2×A2+X3×A3)···(2)

A1+A2+A3=1

In addition, ABS(x) represents the absolute value of x.

A1 obtained in this way is output to the first multiplier 1113 as a gain coefficient 1119 . In addition, A2 is output to the second multiplier 1114 as a gain coefficient 1120 . In addition, A3 is output to the third multiplier 1115 as a gain coefficient 1121 .

In the first multiplier 1113 , the input signal 1110 is multiplied by the gain coefficient 1119 , and the obtained signal 1122 is output to the adder 713 .

In the second multiplier 1114 , the input signal 1111 is multiplied by the gain coefficient 1120 , and the obtained signal 1123 is output to the adder 713 .

In the third multiplier 1115 , the input signal 1112 is multiplied by the gain coefficient 1121 , and the obtained signal 1124 is output to the adder 713 .

In the adder 713 , the three input signals 1122 , 1123 , and 1124 are added, and the obtained signal is output as the output signal 107 .

As above, with the processing structure of the third embodiment, the frequency characteristic deformation part can be realized by using three phase corrections and three HPFs, and the characteristic closer to the normal HPF than that of the first embodiment can be realized by the frequency characteristic deformation part. Such an effect.

Of course, by increasing the phase correction and the number of HPFs, it is possible to realize characteristics closer to normal HPFs. In addition, by substituting the HPF of this configuration with an LPF, it is possible to adjust the frequency characteristics of the signal so that the frequency characteristics of the signal do not exceed the maximum amplitude of the digital signal in a digital audio signal that has undergone frequency characteristic correction such that the correction of high-frequency components is large. out of shape.

Implementation mode 4.

Fig. 10 is a diagram showing another embodiment of the present invention. In this embodiment, an example is shown in which higher harmonics of low frequencies cut off by a high-pass filter are generated and added. Next, the operation of this embodiment will be described.

The input signal 101 input to the signal processing device of the present invention is branched and sent to the phase correction unit 701 and the HPF 702 .

The phase correction unit 701 does not change the frequency-amplitude characteristic of the input signal, but only corrects the phase characteristic so that it becomes substantially the same as the phase characteristic of the HPF 702, and outputs the obtained signal 703 to the first multiplier 705 and the excessive input estimation unit. 102 and a higher harmonic signal generating unit 2001.

The HPF 702 filters the input signal 101 and outputs the obtained signal 704 to the adder 2003 .

Here, the phase correction unit can be realized by an all-pass filter or sampling delay processing similarly to the first embodiment, so detailed description is omitted.

The harmonic signal generation unit 2001 is composed of a low frequency extraction unit 2004 , a harmonic generation unit 2006 , and a multiplier 2008 .

The low frequency extraction unit 2004 receives the output signal 703 of the phase correction unit 701 to extract the low frequency cut off by the HPF 702 , and outputs the obtained signal 2005 to the harmonic generation unit 2006 . Here, regarding the realization method of the low frequency extraction unit 2004 for extracting the low frequency cut off by the HPF 702, there is the following method: as shown in FIG. A method realized by subtraction; a method configured by LPF 2012 as shown in FIG. 12 and realized by passing the output signal 703 of the phase correction unit 701 through a filter having the same filter specification as that of the HPF 702 .

In harmonic generation unit 2006 , harmonics of output signal 2005 of low frequency extraction unit 2004 are generated up to nth order (n is an integer equal to or larger than 3), and the obtained signal 2007 is output to multiplier 2008 . Here, the realization method of the higher harmonic generation unit 2006 is generated by waveform deformation such as peak hold, full-wave rectification, and half-wave rectification, m-time multiplication of the signal 2005 (m is an integer), frequency division, etc., to generate Both odd-numbered harmonics and even-numbered harmonics are sufficient.

The multiplier 2008 multiplies the output signal 2007 of the harmonic generator 2006 by a user-preferred gain factor, and outputs the obtained signal 2002 to the adder 2003 . Here, as for the gain coefficient to be multiplied by the multiplier 2008, a plurality of fixed gain coefficients are prepared in advance, and are changed according to the user's preference.

The adder 2003 adds the two input signals 704 and 2002 , and outputs the obtained signal 2003 ′ to the excessive input estimation unit 102 and the second multiplier 706 .

The excessive input estimation unit 102 of this embodiment is constituted by a speaker diaphragm displacement estimation unit 501 . The speaker diaphragm displacement estimation unit 501 estimates the speaker diaphragm displacement value when the signal 703 is reproduced using information 502 such as the volume value and F0 of the target speaker, and obtains the first speaker diaphragm displacement value 707 . Similarly, the displacement value of the speaker diaphragm when the signal 2003' is reproduced is estimated, and the second speaker diaphragm displacement value 708 is obtained.

As a specific example of the displacement value estimation, it is obtained by the same method as in the first embodiment, so detailed description is omitted.

The two speaker diaphragm displacement values 707 and 708 obtained by the speaker diaphragm displacement estimation unit 501 are output to the control unit 105 .

In the control unit 105, when the two input speaker diaphragm displacement values 707 and 708 are multiplied by different gain coefficients and then added, the absolute value of the amplitude value converges within a predetermined threshold value. a gain factor. However, the total of the two gain coefficients is set to 1.

If the two gain coefficients are changed under such conditions, different low-frequency attenuation effects can be achieved. A specific example of the low-frequency attenuation effect has the same characteristics as in the first embodiment, so detailed description is omitted.

In addition, it is possible to change the addition amount of the cut-off low-frequency harmonics according to the low-frequency attenuation effect. FIG. 13 shows the low-frequency attenuation effect when the HPF 702 is realized with a 2-stage 2-order IIR filter with a cutoff frequency of 80 Hz, and the phase correction unit 701 is realized with a 1-stage 2-order IIR all-pass filter with a cutoff frequency of 80 Hz. Plot of the summed image of the corresponding higher harmonics.

In FIG. 13 , when the gain coefficient for the speaker diaphragm displacement value 707 is set to A1, and the gain coefficient for the speaker diaphragm displacement value 708 is set to A2, 2021 represents a characteristic image of A1=0.0 and A2=1.0, 2022 represents the characteristic image of A1=0.1, A2=0.9, and 2023 represents the characteristic image of A1=0.5, A2=0.5. From this, it can be seen that the larger the value of A2 and the larger the attenuation amount of the low frequency, the more the high harmonics of the low frequency below 80 Hz are added.

In addition, since the specific calculation method of the gain coefficients A1 and A2 is obtained in the same manner as in the first embodiment, description thereof will be omitted.

A1 obtained by the control unit 105 is output to the first multiplier 705 as a gain coefficient 709 . In addition, A2 is output to the second multiplier 706 as a gain coefficient 710 .

In the first multiplier 705 , the input signal 703 is multiplied by the gain coefficient 709 , and the obtained signal 711 is output to the adder 713 .

In the second multiplier 706 , the input signal 2003 ′ is multiplied by the gain coefficient 710 , and the obtained signal 712 is output to the adder 713 .

The adder 713 adds the two input signals 711 and 712 and outputs the obtained signal as the output signal 107 .

As described above, with the processing structure of the fourth embodiment, the harmonic signal generation unit generates the low-frequency harmonics cut off by the frequency characteristic deformation unit up to the nth order (n is an integer greater than or equal to 3), and it is possible to obtain Such an effect of the cut low frequency can be virtually felt through the sound psychological characteristic "Missing fundamental". In addition, since the output signal of the HPF and the gain of the generated harmonic signal are controlled together in the control unit, the low-frequency interpolation effect can also be changed according to the attenuation characteristics of low frequencies.

Here, Missing fundamental refers to a characteristic that, when listening to sounds of two or more frequencies, it is falsely heard as a sound of a different frequency.

Implementation mode 5.

In Embodiment 5, the frequency characteristic deformation part is realized by one phase correction part and one HPF, but the present invention is not limited to this, and the frequency characteristic deformation part may be realized by a plurality of phase correction parts and a plurality of HPFs.

FIG. 14 is a diagram showing an example in which a frequency characteristic deformer is realized by three phase correctors and three HPFs. Next, the operation of this embodiment will be described.

The input signal 101 input to the signal processing device of the present invention is branched into six and sent to the first HPF1101, the second HPF1102, the third HPF1103, the first LPF2101, the second LPF2102, and the third LPF2103.

In the first HPF 1101 , the input signal is filtered, and the obtained signal 1104 is output to the first phase correction unit 1107 .

In the second HPF 1102 , the input signal is filtered, and the obtained signal 1105 is output to the second phase correction unit 1108 .

In the third HPF 1103 , the input signal is filtered, and the obtained signal 1106 is output to the third phase correction unit 1109 .

In the first phase correction unit 1107, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the second HPF 1102 and the third HPF 1103 are processed. The resulting signal 1110 is output to an adder 2125 .

In the second phase correction unit 1108, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the first HPF 1101 and the third HPF 1103 are processed. The obtained signal 1111 is output to the adder 2126 .

In the third phase correction unit 1109, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the first HPF 1101 and the second HPF 1102 are processed. The resulting signal 1112 is output to an adder 2127 .

Here, each phase correction unit can be realized by an all-pass filter or sampling delay processing similarly to the first embodiment, and thus detailed description thereof will be omitted.

In the first LPF 2101 , the input signal 101 is processed by a filter having the same filter specification as that of the first HPF 1101 , and the obtained signal 2104 is output to the fourth phase correction unit 2107 .

In the second LPF 2102 , the input signal 101 is processed by a filter having the same filter specification as that of the second HPF 1102 , and the obtained signal 2105 is output to the fifth phase correction unit 2108 .

In the third LPF 2103 , the input signal 101 is processed by a filter having the same filter specifications as the third HPF 1103 , and the obtained signal 2106 is output to the sixth phase correction unit 2109 .

In the fourth phase correction unit 2107, without changing the frequency-amplitude characteristic of the signal, only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the second LPF 2102 and the third LPF 2103 are processed. The obtained signal 2110 is output to the first harmonic generator 2113 .

In the fifth phase correcting unit 2108, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the first LPF 2101 and the third LPF 2103 are processed. The obtained signal 2111 is output to the second harmonic generator 2114 .

In the sixth phase correction unit 2109, the frequency-amplitude characteristic of the signal is not changed, but only the phase characteristic is corrected so that it becomes substantially the same as the phase characteristic when both the first LPF 2101 and the second LPF 2102 are processed. The obtained signal 2112 is output to the third harmonic generator 2115 .

In the first harmonic generation unit 2113 , the harmonics of the signal 2110 are generated, and the obtained signal 2116 is output to the multiplier 2119 .

In the second harmonic generation unit 2114 , the harmonic of the signal 2111 is generated, and the obtained signal 2117 is output to the multiplier 2120 .

In the third harmonic generation unit 2115 , the harmonic of the signal 2112 is generated, and the obtained signal 2118 is output to the multiplier 2121 .

Here, as for the realization method of each harmonic generation unit, it is generated by waveform deformation such as peak hold, full-wave rectification, and half-wave rectification, m-time multiplication of signals (m is an integer), frequency division, etc., and an odd number is generated. Both sub-harmonics and even-numbered harmonics are sufficient.

In the multiplier 2119 , the signal 2116 is multiplied by a gain factor preferred by the user, and the obtained signal 2122 is output to the adder 2125 .

In the multiplier 2120 , the signal 2117 is multiplied by a gain factor preferred by the user, and the obtained signal 2123 is output to the adder 2126 .

In the multiplier 2121 , the signal 2118 is multiplied by a gain factor preferred by the user, and the obtained signal 2124 is output to the adder 2127 .

Here, for the gain coefficients multiplied by the multipliers 2119 , 2120 , and 2121 , a plurality of fixed gain coefficients are prepared in advance and changed according to the user's preference.

The adder 2125 adds the two input signals 1110 and 2122 and outputs the obtained signal 2128 to the excessive input estimation unit 102 and the first multiplier 1113 .

The adder 2126 adds the two input signals 1111 and 2123 and outputs the obtained signal 2129 to the excessive input estimation unit 102 and the second multiplier 1114 .

The adder 2127 adds the two input signals 1112 and 2124 , and outputs the obtained signal 2130 to the excessive input estimation unit 102 and the third multiplier 1115 .

The excessive input estimation unit 102 is composed of a speaker diaphragm displacement estimation unit 501 . The speaker diaphragm displacement estimation unit 501 estimates the speaker diaphragm displacement value when the signal 2128 is reproduced using the information 502 such as the volume value and F0 of the target speaker, and obtains the first speaker diaphragm displacement value 1116 . Similarly, the displacement value of the speaker diaphragm when the signal 2129 is reproduced is estimated, and the second speaker diaphragm displacement value 1117 is obtained. Similarly, the displacement value of the speaker diaphragm when the signal 2130 is reproduced is estimated, and the third speaker diaphragm displacement value 1118 is obtained.

As a specific example of the displacement value estimation, it is obtained by the same method as in the first embodiment, so detailed description is omitted.

The three speaker diaphragm displacement values 1116 , 1117 , and 1118 calculated by the speaker diaphragm displacement estimation unit 501 are output to the control unit 105 .

In the control unit 105, when the three input speaker diaphragm displacement values 1116, 1117, and 1118 are multiplied by different gain coefficients and added, the absolute value of the amplitude value converges within a predetermined threshold value. The 3 gain coefficients. However, the total of the three gain coefficients is set to 1.

If the three gain coefficients are changed under such conditions, different low-frequency attenuation effects can be achieved. A specific example of the low-frequency attenuation effect has the same characteristics as in the third embodiment, so detailed description is omitted.

In addition, it is possible to change the addition amount of the cut-off low-frequency harmonics according to the low-frequency attenuation effect. FIG. 15 is a diagram showing an added image of higher harmonics corresponding to the low-frequency attenuation effect in the case where the first HPF 1101 is realized by a two-stage second-order IIR filter with a cutoff frequency of 30 Hz. , using a 2-stage 2-order IIR filter with a cut-off frequency of 70Hz to realize the second HPF1102, using a 4-stage 2-order IIR filter with a cut-off frequency of 140Hz to realize the third HPF1103, and using a 1-stage 2-order IIR filter with a cut-off frequency of 70Hz The filter and the 2-stage 2nd-order IIR filter with a cutoff frequency of 140 Hz are connected in series to realize the first phase correction unit 1107, and the 1st-stage 2nd-order IIR filter with a cutoff frequency of 30 Hz and the 2-stage 2nd-order IIR filter with a cutoff frequency of 140 Hz are connected in series. The second-order IIR filter is connected in series to realize the second phase correction unit 1108, and the first-stage second-order IIR filter with a cutoff frequency of 30 Hz and the first-stage second-order IIR filter with a cutoff frequency of 70 Hz are connected in series to realize the second phase correction unit 1108. Three-phase correction unit 1109 .

In Fig. 15, the gain coefficient for the speaker diaphragm displacement value 1116 is set as A1, the gain coefficient for the speaker diaphragm displacement value 1117 is set as A2, and the gain coefficient for the speaker diaphragm displacement value 1118 is set as A3 When, 2131 represents the characteristic image of A1=1.0, A2=0.0, A3=0.0, 2132 represents the characteristic image of A1=0.9, A2=0.1, A3=0.0, 2133 represents the characteristic image of A1=0.0, A2=0.0, A3=1.0 Feature image. From this, it can be seen that the low-frequency harmonics cut off by the HPF are added to the frequency band above the cut-off frequency band.

In addition, since the specific calculation method of the gain coefficients A1, A2, and A3 is obtained in the same manner as in the third embodiment, description thereof will be omitted.

A1 obtained by the control unit 105 is output to the first multiplier 1113 as a gain coefficient 1119 . In addition, A2 is output to the second multiplier 1114 as a gain coefficient 1120 . In addition, A3 is output to the third multiplier 1115 as a gain coefficient 1121 .

In the first multiplier 1113 , the input signal 2128 is multiplied by the gain coefficient 1119 , and the obtained signal 1122 is output to the adder 713 .

In the second multiplier 1114 , the input signal 2129 is multiplied by the gain coefficient 1120 , and the obtained signal 1123 is output to the adder 713 .

In the third multiplier 1115 , the input signal 2130 is multiplied by the gain coefficient 1121 , and the obtained signal 1124 is output to the adder 713 .

The adder 713 adds the three input signals 1122 , 1123 , and 1124 , and outputs the obtained signal as the output signal 107 .

As described above, with the processing structure of the fifth embodiment, the harmonic signal generation unit generates the low-frequency harmonics cut off by the frequency characteristic deformation unit up to the nth order (n is an integer greater than or equal to 3), so that The effect of being able to virtually feel the cut low frequency through the psychological characteristic of sound "Missing fundamental" is obtained. In addition, since the output signal of the HPF and the gain of the generated harmonic signal are controlled together in the control unit, the low-frequency interpolation effect can also be changed according to the attenuation characteristics of low frequencies.

Embodiment 6

In Embodiments 4 and 5, the output signal 2002 of the harmonic signal generation unit 2001 is added to the output signal of the HPF, but it may be configured to add the harmonic signal at the stage after the frequency characteristic deformation unit 103. The output signal 2002 of the wave signal generator 2001.

FIG. 16 is a diagram showing an example in which the addition position of the output signal 2002 of the harmonic signal generation unit 2001 is changed to the configuration of the subsequent stage of the frequency characteristic deformation unit 103 in the fourth embodiment. Next, the operation of this embodiment will be described.

The input signal 101 input to the signal processing device of the present invention is branched and sent to the phase correction unit 701 and the HPF 702 .

The phase correction unit 701 does not change the frequency-amplitude characteristic of the input signal, but only corrects the phase characteristic so that it becomes substantially the same as the phase characteristic of the HPF 702, and outputs the obtained signal 703 to the first multiplier 705 and the excessive input estimation unit. 102 and a higher harmonic signal generating unit 2001.

The HPF 702 filters the input signal 101 and outputs the obtained signal 704 to the multiplier 706 and the excessive input estimation unit 102 .

Here, the phase correction unit can be realized by an all-pass filter or sampling delay processing similarly to the first embodiment, so detailed description is omitted.

The excessive input estimation unit 102 of this embodiment is constituted by a speaker diaphragm displacement estimation unit 501 . The speaker diaphragm displacement estimation unit 501 estimates the speaker diaphragm displacement value when the signal 703 is reproduced using information 502 such as the volume value and F0 of the target speaker, and obtains the first speaker diaphragm displacement value 707 . Similarly, the displacement value of the speaker diaphragm when the signal 704 is reproduced is estimated, and the second speaker diaphragm displacement value 708 is obtained.

As a specific example of the displacement value estimation, it is obtained by the same method as in the first embodiment, so detailed description is omitted.

The two speaker diaphragm displacement values 707 and 708 obtained by the speaker diaphragm displacement estimation unit 501 are output to the control unit 105 .

In the control unit 105, when the two input speaker diaphragm displacement values 707 and 708 are multiplied by different gain coefficients and then added, the absolute value of the amplitude value converges within a predetermined threshold value. a gain factor. However, the total of the two gain coefficients is set to 1.

If the two gain coefficients are changed under such conditions, different low-frequency attenuation effects can be achieved. A specific example of the low-frequency attenuation effect has the same characteristics as in the first embodiment, so detailed description is omitted.

In addition, in FIG. 16 , when the gain coefficient for the speaker diaphragm displacement value 707 is set to A1, and the gain coefficient for the speaker diaphragm displacement value 708 is set to A2, the specific calculation method of the gain coefficients A1 and A2 , was obtained in the same manner as in Example 1, and thus description thereof is omitted.

A1 obtained by the control unit 105 is output to the first multiplier 705 as a gain coefficient 709 . In addition, A2 is output to the second multiplier 706 and the multiplier 2201 as the gain coefficient 710 .

In the first multiplier 705 , the input signal 703 is multiplied by the gain coefficient 709 , and the obtained signal 711 is output to the adder 713 .

In the second multiplier 706 , the input signal 704 is multiplied by the gain coefficient 710 , and the obtained signal 712 is output to the adder 713 .

The adder 713 adds the two input signals 711 and 712 , and outputs the obtained signal 2203 to the adder 2204 .

The harmonic signal generation unit 2001 is composed of a low frequency extraction unit 2004 , a harmonic generation unit 2006 and a multiplier 2008 .

The output signal of the phase correction unit 701 is input to the low frequency extraction unit 2004 , the low frequency cut off by the HPF 702 is extracted, and the obtained signal 2005 is output to the harmonic generation unit 2006 . Here, the implementation method of the low-frequency extraction unit 2004 for extracting the low-frequency cut off by the HPF 702 is obtained in the same manner as in the fourth embodiment, so the description thereof will be omitted.

In harmonic generation unit 2006 , harmonics of output signal 2005 of low frequency extraction unit 2004 are generated up to nth order (n is an integer equal to or larger than 3), and the obtained signal 2007 is output to multiplier 2008 . Here, the realization method of the higher harmonic generation unit 2006 is generated by waveform deformation such as peak hold, full-wave rectification, and half-wave rectification, m-time multiplication of the signal 2005 (m is an integer), frequency division, etc., to generate Both odd-numbered harmonics and even-numbered harmonics are sufficient.

The multiplier 2008 multiplies the output signal 2007 of the harmonic generator 2006 by a user-preferred gain factor, and outputs the obtained signal 2002 to the multiplier 2201 . Here, as for the gain coefficient multiplied by the multiplier 2008, a plurality of fixed gain coefficients are prepared in advance, and are changed according to the user's preference.

In the multiplier 2201 , the input signal 2002 is multiplied by the gain coefficient 710 , and the obtained signal 2202 is output to the adder 2204 . Here, in the multiplier 2201, it is possible to change the addition amount of the cut-off low-frequency higher harmonics according to the low-frequency attenuation effect. The addition image of the higher harmonics corresponding to the low-frequency attenuation effect is the same as that of Embodiment 4, and therefore description thereof will be omitted.

The adder 2204 adds the two input signals 2202 and 2203 and outputs the obtained signal as the output signal 107 .

As described above, with the processing structure of the sixth embodiment, the harmonic signal generation unit generates the low-frequency harmonics cut off by the frequency characteristic deformation unit up to the nth order (n is an integer greater than or equal to 3), so that It is possible to obtain the effect of virtually feeling the cut low frequency through the "Missing fundamental" psychological characteristic of the sound. In addition, since the output signal of the HPF and the gain of the generated harmonic signal are controlled together in the control unit, the low-frequency interpolation effect can also be changed according to the attenuation characteristics of low frequencies.

In addition, the invention of the present application can realize free combination of each embodiment, modification of arbitrary components of each embodiment, or omission of arbitrary components of each embodiment within the scope of the present invention.

Industrial availability

As described above, the frequency characteristic deforming device of the present invention can improve distortion and broken sound during audio signal reproduction, and can be used in audio reproduction devices and the like.

Claims (11)

1. A frequency characteristic distortion device, is characterized in that, possesses: a filter that deforms frequency characteristics of a signal as an object; a phase correction unit that corrects a phase characteristic of the target signal so as to be substantially the same as a phase characteristic of the filter; a first multiplier that adjusts the gain of the signal output from the phase correction section; a second multiplier to adjust the gain of the signal output from said filter; a coefficient determination unit that determines gain coefficients of the first and second multipliers such that the sum of the gain coefficients of the first multiplier and the gain coefficients of the second multiplier becomes a fixed value; and The adder adds the two signals output from the first multiplier and the second multiplier. 2. frequency characteristic deformation device according to claim 1, is characterized in that, The filter consists of a high-pass filter, The coefficient determination unit brings the gain coefficient of the second multiplier closer to the fixed value when the cutoff frequency of the target characteristic is increased, and the coefficient is adjusted when the cutoff frequency of the target characteristic is lowered. The determining unit makes the gain coefficient of the second multiplier close to 0 to change the cutoff frequency of the target characteristic. 3. frequency characteristic distortion device according to claim 1, is characterized in that, The filter consists of a low-pass filter, The coefficient determination unit makes the gain coefficient of the second multiplier close to 0 when the cutoff frequency of the target characteristic is increased, and the coefficient determination unit makes the gain coefficient of the second multiplier close to 0 when the cutoff frequency of the target characteristic is lowered. The gain coefficient of the second multiplier is close to the fixed value, so that the cutoff frequency of the target characteristic is changed. 4. frequency characteristic distortion device according to claim 2, is characterized in that, possesses: Multiple high-pass filters with different cutoff frequencies; a phase correction unit that corrects the phase characteristics of the target signal so as to be substantially the same as the phase characteristics of each of the high-pass filters; a plurality of multipliers that adjust gains of signals output from the high-pass filter and the phase correction section; and a coefficient determining unit that determines each gain coefficient so that the total of the gain coefficients of the plurality of multipliers becomes a fixed value, The frequency characteristic deforming device generates a plurality of filtered output signals by passing the target signal through the plurality of high-pass filters, and corrects each generated filter output signal using a phase correcting unit having a phase characteristic equivalent to another high-pass filter. The phase characteristic of the output signal is set to be approximately the same as the phase characteristic of each filtered output signal, and when the cutoff frequency of the characteristic as the target is lowered, the output signal from the filter with a lower cutoff frequency is compared with the output signal by the coefficient determination unit. The gain coefficient of the corresponding multiplier is close to the fixed value, and when the cutoff frequency of the target characteristic is increased, the gain of the multiplier corresponding to the output signal from the filter with a high cutoff frequency is adjusted by the coefficient determination unit. The coefficients are close to the fixed value, and after weighting the phase-corrected filtered output signals by the gain coefficients determined by the coefficient determination unit, the signals are added by the adder, so that The cutoff frequency of the characteristic of the target changes. 5. frequency characteristic deformation device according to claim 3, is characterized in that, possesses: Multiple low-pass filters with different cutoff frequencies; a phase correction unit that corrects the phase characteristics of the target signal to be substantially the same as the phase characteristics of each of the low-pass filters; a plurality of multipliers that adjust gains of signals output from the low-pass filter and the phase correction section; and a coefficient determining unit that determines each gain coefficient so that the total of the gain coefficients of the plurality of multipliers becomes a fixed value, The frequency characteristic deforming device generates a plurality of filtered output signals by passing the target signal through the plurality of low-pass filters, and corrects the generated output signals using a phase correction unit having a phase characteristic equivalent to that of other low-pass filters. The phase characteristics of the respective filter output signals are set to be approximately the same, and when the cutoff frequency of the target characteristic is lowered, the coefficient determination unit is used to compare the phase characteristics of the filter output signals with the low cutoff frequency. The gain coefficient of the multiplier corresponding to the output signal is close to the fixed value, and when the cutoff frequency of the target characteristic is increased, the multiplier corresponding to the output signal from the filter with a high cutoff frequency is adjusted by the coefficient determination unit. The gain coefficient is close to the fixed value, and after weighting the phase-corrected filtered output signals by the gain coefficients determined by the coefficient determination unit, the signals are added by the adder, thereby Change the cutoff frequency of the target characteristic. 6. frequency characteristic deformation device according to claim 2, is characterized in that, possesses: a low-frequency extracting section extracting low frequencies that would be cut off by the high-pass filter; and The harmonic generation unit generates harmonics of the signal output from the low frequency extraction unit. 7. frequency characteristic distortion device according to claim 4, is characterized in that, possesses: a low-frequency extracting section extracting low frequencies that would be cut off by the high-pass filter; and The harmonic generation unit generates harmonics of the signal output from the low frequency extraction unit. 8. frequency characteristic distortion device according to claim 2, is characterized in that, possesses: a low-frequency extracting unit for extracting low frequencies that will be cut off by the high-pass filter; a harmonic generation unit that generates harmonics of the signal output from the low frequency extraction unit; and an adder that adds the signal output from the higher harmonic generation unit to the signal output from the high-pass filter, In the second multiplier, the gain of the signal output from the adder is changed, and the gain of the signal output from the harmonic generator is changed according to the attenuation characteristics of low frequencies. 9. frequency characteristic deformation device according to claim 2, is characterized in that, possesses: a low-frequency extracting unit for extracting low frequencies that will be cut off by the high-pass filter; a harmonic generation unit that generates harmonics of the signal output from the low frequency extraction unit; and a multiplier for multiplying the signal output from the higher harmonic generation unit by a gain factor multiplied by the second multiplier, The gain of the signal output from the harmonic generator is changed according to the attenuation characteristic of the low frequency. 10. frequency characteristic distortion device according to claim 4, is characterized in that, possesses: a plurality of low-frequency extractors for extracting low frequencies cut off by the high-pass filter; a plurality of harmonic generators generating harmonics of signals output from the plurality of low frequency extractors; and a plurality of adders for adding a plurality of signals output from the higher harmonic generation unit and a plurality of signals output from the high-pass filter, In the plurality of multipliers according to claim 4, the gains of the plurality of signals output from the adder are changed, and the gains of the plurality of signals output from the high-order harmonic generation unit are changed according to the attenuation characteristics of low frequencies. gain. 11. frequency characteristic deformation device according to claim 4, is characterized in that, possesses: a plurality of low-frequency extractors for extracting low frequencies cut off by the high-pass filter; a plurality of harmonic generators generating harmonics of signals output from the plurality of low frequency extractors; and a plurality of multipliers for multiplying the signals output from the plurality of higher harmonic generators by a plurality of gain coefficients multiplied in the multiplier according to claim 4 , Gains of the plurality of signals output from the harmonic generator are changed according to attenuation characteristics of low frequencies.